Mobile robots play an important role in human-robot interaction and they can be regarded as one of the major focus researches conducted in TaarLab. Mobile robots are usually classified based on the following criteria:
-The environment in which they travel such as Unmanned Ground Vehicles (UGVs).
-The device they use to move, mainly such as legged robot and wheeled robot .
In TaarLab, researches on mobile robots due its great importance in human-robot interaction application cover bothpractical and theoretical aspects where more emphasis is placed on wheeled robot. Various projects can be defined on a wheeled robot and TaarLab for its practical purposes needs a robot with particular characteristics, such as proximity sensor, camera, accelerometer, sound sensor and etc. In this category, TaarLab is equipped with 7 e-pucks which most of the proposed algorithms are tested and examined using these e-pucks. Recently, TaarLab launched a project in order to build FPGA-based mobile robots. For more information regarding our FPGA-based mobile robot click here. Projects defined on this regard, which are essentially based on e-pucks, can be classified as follows:
The simultaneous localisation and map building (SLAM) problem asks if it is possible for an autonomous vehicle to start in an unknown location in an unknown environment and then to incrementally build a map of this environment while simultaneously using this map to compute absolute vehicle location . SLAM usually needs high precision ranging data, for example, a 2D laser range scanner. SLAM project in TaarLab, uses an implicit way of forming a representation of the robot’s environment that is based on a Neural Network, for example, classic methods such as MLP network, or neuro-fuzzy methods such as ANFIS network.
Navigation using MPC
Model Predictive Control, or MPC, is an advanced method of process control that has been in use in the process. Model predictive controllers rely on dynamic models of the process, most often linear empirical models obtained by system identification . By using model predictive control (MPC), a discontinuous control law is naturally obtained. One of the main advantages of MPC is the ability to handle constraints (due to state or input limitations) in a straightforward way . MPC is implemented for navigation in unknown environment and optimization is solved using convex concept after prediction of system states.
Path planning is one of the important parts in mobile robot filed. Path planning is done in known environment and is usually implemented using offline methods such as optimal control. Path planning is usually designed base on consuming minimum energy and minimum waste time. Path planning in Taar, is carried out using optimal control method and implicit method “interval analysis” to design a safe way for passing robots.
Obstacle avoidance is the task of satisfying some control objective subject to non-intersection or non-collision position constraints . For the mobile robot navigation, the potential field method is popular, because it can unify path planning, trajectory and control into one problem. The basic concept of the potential field method is to fill the robot’s workspace with an artificial potential field in which the robot is attracted to its target position and is repulsed away from the obstacles . In Taar, obstacle avoidance is implemented using some methods such as potential field, bug algorithm and etc.
After path planning, path tracking is important issue which robot must track the optimal path with minimum error. The two wheeled mobile robot system has two inputs (translational velocity and angular velocity) and three outputs (centre positions x, y and heading angle of the mobile robot on two dimensional Cartesian workspace). The controller can track the desired path by using these inputs and feedback of the output. Designing a proper controller lead to path tracking. In TaarLab, various controllers for path tracking was designed, for example, control using optimal control method or using nonlinear control concept such as backstepping method.
Detection of environment using search algorithm is one of the some projects which is carried out in TaarLab. Searching important objects such as finding a person has been injured in a building destroyed, is the main purpose of this project. Detection of environment is done with camera and image processing on the photo. Detection of objects in unknown environment was implemented in TaarLab last autumn. E-puck has detected the red balls using the search algorithm based on Closeness of the ball and the results of this project were submitted in multibody dynamic conference. In this project, when the robot see the target, the robot rotates in such a way that it would be exactly placed in the midline of the object by using the results obtained by the image processing. Then the robot moves toward the nearest object until the robot is enough close to the object. The position of the robot is determined by odometry method and the position of object is computed by the position of robot, accordingly.
- Building a mobile robot based on Embedded system concept
An important project in mobile robot part of TaarLab, is building a mobile robot with high precision technology and tools. This mobile robot is designed on the new method of building of mobile robot and is built using Field-Programmable Gate Array (FPGA).
 Dissanayake, MWM Gamini, et al. “A solution to the simultaneous localization and map building (SLAM) problem.” Robotics and Automation, IEEE Transactions on 17.3 (2001): 229-241.
 Kuhne, Felipe, Walter Fetter Lages, and Joao Manoel Gomes da Silva Jr. “Model predictive control of a mobile robot using linearization.” Proceedings of Mechatronics and Robotics. 2004.
 Jiang, Lihua, Mingcong Deng, and Akira Inoue. “Obstacle avoidance and motion control of a two wheeled mobile robot using SVR technique.” International Journal of Innovative Computing, Information and Control 5.2 (2009): 253-262.